The Prevalence and Clinical Significance of Anaerobic Bacteria

antibiotics

Article The Prevalence and Clinical Signiﬁcance of Anaerobic Bacteria in Major Liver Resection

Jens Strohäker *, Sophia Bareiß, Silvio Nadalin, Alfred Königsrainer , Ruth Ladurner and Anke Meier

Department of General, Visceral and Transplantation Surgery, University Hospital of Tuebingen, 72076 Tuebingen, Germany; [email protected] (S.B.); [email protected] (S.N.); [email protected] (A.K.); [email protected] (R.L.); [email protected] (A.M.) * Correspondence: [email protected]; Tel.: +49-7071-29-68171

Abstract: (1) Background: Anaerobic infections in hepatobiliary surgery have rarely been addressed. Whereas infectious complications during the perioperative phase of liver resections are common, there are very limited data on the prevalence and clinical role of anaerobes in this context. Given the risk of contaminated bile in liver resections, the goal of our study was to investigate the prevalence and outcome of anaerobic infections in major hepatectomies. (2) Methods: We retrospectively analyzed the charts of 245 consecutive major hepatectomies that were performed at the department of General, Visceral, and Transplantation Surgery of the University Hospital of Tuebingen between July 2017 and August 2020. All microbiological cultures were screened for the prevalence of anaerobic bacteria and the patients’ clinical characteristics and outcomes were evaluated. (3) Results: Of the 245 patients, 13 patients suffered from anaerobic infections. Seven had positive cultures from the biliary tract during the primary procedure, while six had positive culture results from samples obtained during the management of complications. Risk factors for anaerobic infections were preoperative biliary stenting (p = 0.002) and bile leaks (p = 0.009). All of these infections had to be treated by Citation: Strohäker, J.; Bareiß, S.; intervention and adjunct antibiotic treatment with broad spectrum antibiotics. (4) Conclusions: Nadalin, S.; Königsrainer, A.; Ladurner, Anaerobic infections are rare in liver resections. Certain risk factors trigger the antibiotic coverage R.; Meier, A. The Prevalence and of anaerobes. Clinical Signiﬁcance of Anaerobic Bacteria in Major Liver Resection. Keywords: anaerobic infection; liver resection; cholangitis; biliary tract infection Antibiotics 2021, 10, 139. https:// doi.org/10.3390/antibiotics10020139

Academic Editor: Fernando Cobo 1. Introduction Received: 6 January 2021 Accepted: 28 January 2021 Liver resections are widely available procedures to cure benign and malignant diseases Published: 31 January 2021 of the liver. After partial hepatectomy bile leaks and surgical site infections, complications are feared, since they drastically increase morbidity and mortality [1–3]. Liver resections Publisher’s Note: MDPI stays neutral are considered clean-contaminated procedures given the risk of preexisting bacterial and with regard to jurisdictional claims in fungal colonization of the biliary system. The most common bacteria cultured from the bile published maps and institutional afﬁl- duct and postoperative infectious complications are of gastrointestinal origin [4–6]. iations. Anaerobic liver infections, however, are rare in hepatobiliary surgery. The most commonly described presentation of anaerobes in the liver is in the form of liver abscesses. These abscesses are frequently caused by acute or chronic abdominal inﬂammation or infection [7,8]. Recently, an emerging number of abscesses have been of iatrogenic origin,

Copyright: © 2021 by the authors. caused by percutaneous ablation of primary and secondary liver malignancies. Abscesses Licensee MDPI, Basel, Switzerland. after ablation appear to develop in up to 2% of patients [9,10]. Anaerobic bacteria are com- This article is an open access article mon pathogens in the intestines. The liver and native biliary tract are generally uncommon distributed under the terms and habitats for anaerobic bacteria. However, they may be cultured from the biliary tract in up conditions of the Creative Commons to 20% of patients in the presence of a biliary tract occlusion or stent [11,12]. Aside from Attribution (CC BY) license (https:// biliary stents, bilioenteric anastomoses are considered a risk factor for anaerobic coloniza- creativecommons.org/licenses/by/ tion and infection of the biliary tree (and liver). Therefore, the 2018 Tokyo Guidelines 4.0/).

Antibiotics 2021, 10, 139. https://doi.org/10.3390/antibiotics10020139 https://www.mdpi.com/journal/antibiotics Antibiotics 2021, 10, 139 2 of 12

(TG2018) recommend covering anaerobic bacteria when treating cholangitis in the presence of bilioenteric anastomosis [13]. The department of General, Visceral, and Transplantation Surgery of the Tuebingen University Hospital is a Tertiary Care Academic Teaching facility that performs an av- erage of ~200 liver resections as well as ~50 liver transplantations per year. Our center is specialized on complex liver resections with a clinical and scientific focus on biliary malignancies. [14,15]. A major hepatectomy/liver resection is defined as the removal of ≥3 liver segments [16]. Due to biliary obstruction of perihilar malignancies, patients often present with jaundice or have undergone preoperative stenting and are thus at risk of biliary tract infections [12]. Infectious complications after surgery are still a major risk factor for morbidity and mortality after liver resection. Surgical site and organ-space infections present in up to 20% of major hepatectomies [1,17]. The role of anaerobic bacteria is yet to be assessed. To our knowledge, there are neither sufficient data on the presence of anaerobes in intraoperative microbiological specimen from liver resections nor from surgical site infection (SSI) after hepatectomy. Even less is known about the presence of anaerobic bacteria’s antibiotic resistance to commonly used antibiotics after liver surgery. The goal of this study was to evaluate the prevalence and clinical role of anaerobic bacteria in patients undergoing major liver resection for benign and malignant disease.

2. Results 2.1. Clinical Characteristics We analyzed the charts of all consecutive patients that had undergone major liver resection at the department of General, Visceral, and Transplantation Surgery of the Tuebingen University Hospital from June 2017 to August 2020. We included all adult (age ≥ 16 years) consecutive patients that underwent laparotomy and had at least three liver segments removed. Pediatric patients, as well as patients that had minor liver resections, were ex- cluded. Based on the preoperative diagnosis and intraoperative ﬁndings, material was sent for microbiological testing at the attending surgeon’s discretion. During the study period, 245 patients met the inclusion criteria. Of these 245 patients, 76 had intraoperative material sent for microbiological testing, of which 49 showed growth of pathogenic bacteria. Furthermore, 102 of the 245 patients had culture-proven microbiological growth during the ﬁrst 30 postoperative days (this includes the 49 that had growth on the intraoperative cultures). From these 102 patients, we were able to identify 13 patients that had anaerobic infections. Of these 13 patients, 7 patients had positive anaerobic cultures from specimens that were collected during under the primary surgical procedure. The remaining six patients had positive anaerobic cultures during revision (n = 5) procedures or from specimen collected from drains (n = 1). The median age was 60 years (Standard Deviation (SD) ±15, range 34–80). Five patients were male (38%), and eight were female (62%). Most patients underwent liver resection for malignancy (n = 9), while the other patients were treated for helminthic disease (n = 2), suspected cholangiocarcinoma (n = 1), and recurrent cholangitis after cholecystectomy (n = 1). Of the 13 patients, 9 patients received prolonged perioperative antibiotics for suspected contaminated biliary tract (n = 3) or risk of post-hepatectomy liver failure (PHLF) (n = 6). These patients had anaerobic coverage with metronidazole (n = 6) or piperacillin/tazobactam (n = 2) or meropenem (n = 1). For details, see Table1. Antibiotics 2021, 10, 139 3 of 12

Table 1. The surgical procedures and outcomes, as well as the isolated anaerobic strains of the 13 patients that had positive anaerobic cultures. * Procedures according to Brisbane classification.

Surgical Details and Outcome Location of Preoperative Age, Gender Diagnosis Procedure Anaerobe Complications Length of Stay Outcome Specimen Stent Right Trisectionectomy * Prevotella buccae + Resection of Bile duct 34 f iCCA Prevotella Yes 11d alive Extrahepatic bile duct and 1st Procedure melaninogenica bilioenteric anastomosis Right Trisectionectomy death from septic Anaerobe not + Resection of Bile duct shock due to 72 m HCC otherwise Yes 29d dead Extrahepatic bile duct and 1st Procedure post-hepatectomy speciﬁed bilioenteric anastomosis liver failure death from multi-organ failure with bile Embryonal Right Trisectionectomy Anastomotic leak Bacteroides leak and 44 w Sarcoma of the No 35d dead + Segmental Colectomy Revision vulgatus peritonitis Liver & post-hepatectomy liver failure Right Trisectionectomy Echinococcus + Resection of Bile duct Biﬁdobacterium 38 m Yes 28d alive alveolaris Extrahepatic bile duct and 1st Procedure animalis bilioenteric anastomosis Left Hepatectomy + Resection of Bile duct Bacteroides 71 m phCCA Yes 13d alive Extrahepatic bile duct and 1st Procedure thetaiotaomicron bilioenteric anastomosis Left Hepatectomy + Resection of Bile duct 78 f phCCA Prevotella buccae Yes 20d alive Extrahepatic bile duct and 1st Procedure bilioenteric anastomosis Antibiotics 2021, 10, 139 4 of 12

Table 1. Cont.

Surgical Details and Outcome Location of Preoperative Age, Gender Diagnosis Procedure Anaerobe Complications Length of Stay Outcome Specimen Stent Abscess perihepatic Echinococcus perihepatic and 40 m Right Hepatectomy Finegoldia magna No abscess, wound 29d alive granulosus subcutaneous dehiscence Revision Right Trisectionectomy Klatskin- + Resection of Bilioma Prevotella 80 m mimicking Yes infected bilioma 35d alive Extrahepatic bile duct and CT-Drain melaninogenica Lesion bilioenteric anastomosis Right Trisectionectomy revision Bilophila + Resection of Bile duct laparotomy for 68 f GBCA wadsworthia Yes 31d alive Extrahepatic bile duct and 1st Procedure postoperative Veillonella parvula bilioenteric anastomosis bleeding Right Hepatectomy Recurrent + Resection of Bile duct Bacteroides 54 f No 8d alive Cholangitis Extrahepatic bile duct and 1st Procedure fragilis bilioenteric anastomosis Right Trisectionectomy revision Deep + Resection of laparotomy for 58 f iCCA subcutaneous Prevotella buccae No 25d alive Extrahepatic bile duct and postoperative abscess bilioenteric anastomosis bleeding Right Trisectionectomy Deep revision + Resection of 60 f NET-Metastasis subcutaneous Finegoldia magna No laparotomy for 36d alive Extrahepatic bile duct and abscess postoperative ileus bilioenteric anastomosis revision Bilioma laparotomy for 60 f CRLM Left Hepatectomy Finegoldia magna No 21d alive Revision postoperative bile leak Antibiotics 2021, 10, 139 5 of 12

2.2. Clinical Comparison of Anaerobic Infections We compared the patient group that suffered from anaerobic infections to the control group, which consisted of the remaining 232 patients. Gender distribution (38% male vs. 56% male; p 0.204 X2) was similar in both groups. The same was true for age (60 years standard deviation (SD) ± 15 vs. 64 years SD ± 13; p 0.373 MWU) and body mass index (BMI) (25.2 kg/m2 SD ± 4.7 vs. 24.6 kg/m2 SD ± 4.6; p 0.415 MWU). There were similar rates of malignant (69% vs. 88%; p 0.052 X2) and infectious (8% vs. 15 %; p 0.429 X2) underlying disease. In the anaerobic group, there were signiﬁcantly higher rates of preoperative stents (54% vs. 27%; p 0.000 X2), bile duct resections and reconstructions (69% vs. 32%; p 0.006 X2), and bile leaks (54% vs. 18%; p 0.001 X2). Mean operating time was longer (p 0.035 MWU) in the anaerobic group (likely due to the higher rate of bile duct reconstructions), as well as the length of stay (p 0.002 MWU). The attending surgeon decided whether to activate postoperative preventive measures to preemptively treat post-hepatectomy liver failure in 78% of patients of the anaerobic group and 63% of the control group (p 0.390 X2). For details, see Table2.

Table 2. The comparison of the group of patients that grew anaerobic bacteria and the control group that did not.

Anaerobic Group vs. Control Group Anaerobes No Anaerobes p (n = 13) (n = 232) Gender m (Percentage) 5 (38%) 131 (56%) 0.204 Median age in years 60 (± 15) 64 (± 13) 0.373 Body mass index (BMI) in kg/m2 25.2 (± 4.7) 24.6 (± 4.6) 0.415 Diabetes 3 (23%) 35 (15%) 0.439 Diagnosis Malignant 9 (69%) 204 (88%) 0.052 Recurrent infections 1 (8%) 35 (15%) 0.429 Preoperative biliary stent 7 (54%) 85 (27%) 0.000 Procedure Right Hepatectomy 2 (15%) 91 (39%) 0.114 Right Trisectionectomy 8 (62%) 75 (32%) Left Hepatectomy 3 (23%) 33 (14%) Left Trisectionectomy 0 25 (11%) Central Resection 0 5 (2%) Median operating time in minutes 292 (± 73) 247 (± 100) 0.035 Median length of stay in days 28 (± 9) 13 (± 13) 0.002 Resection of extrahepatic bile duct 9 (69%) 75 (32%) 0.006 Bile leak 7 (54%) 41 (18%) 0.001 Elevated risk for PHLF 10 (78%) 147 (63%) 0.390

2.3. Microbiological Results Of the 245 patients in this study, 102 had positive microbiological cultures throughout their hospital stay and 13 patients grew anaerobic bacteria. Of these 13 patients, 7 patients had cultures from intraoperatively collected bile specimen. The remaining 6 patients had positive cultures from Bilioma (n = 2), perihepatic abscess (n = 1), subcutaneous abscess (n = 2), and enteric anastomotic leak (n = 1). Overall, there were 14 different strains of anaerobic bacteria cultured from these patients, and 1 patient had anaerobic bacteri, that we were unable to identify. Prevotella species was the most commonly isolated anaerobic species (n = 5; 33%), followed by Bacteroides strains (n = 3; 20%) and Finegoldia magna (n = 3; 20%). The latter was only cultured from specimen collected from “complications,” i.e., one bilioma (Figure1), one perihepatic, and one subcutaneous abscess. Except for the perihepatic abscess caused by Finegoldia magna, all specimen that grew anaerobes showed polymicrobial growth of anaerobic and (multiple) aerobic bacteria. For details, see Table3. Antibiotics 2021, 10, x 5 of 12

Elevated risk for PHLF 10 (78%) 147 (63%) 0.390

2.3. Microbiological Results Of the 245 patients in this study, 102 had positive microbiological cultures throughout their hospital stay and 13 patients grew anaerobic bacteria. Of these 13 patients, 7 patients had cultures from intraoperatively collected bile specimen. The remaining 6 patients had positive cultures from Bilioma (n = 2), perihepatic abscess (n = 1), subcutaneous abscess (n = 2), and enteric anastomotic leak (n = 1). Overall, there were 14 different strains of anaerobic bacteria cultured from these patients, and 1 patient had anaerobic bacteri, that we were unable to identify. Prevotella species was the most commonly isolated anaerobic species (n = 5; 33%), followed by Bacteroides strains (n = 3; 20%) and Finegoldia magna (n = 3; 20%). The latter was only cultured from specimen collected from “complications,” i.e., one bilioma (Figure 1), one perihepatic, and one subcutaneous Antibiotics 2021, 10, 139 6 of 12 abscess. Except for the perihepatic abscess caused by Finegoldia magna, all specimen that grew anaerobes showed polymicrobial growth of anaerobic and (multiple) aerobic bacteria. For details, see Table 3.

* RL RL *

Figure 1. The CT scans of a patient that suffered from post-hepatectomy bilioma (white asterisk) originating from the Figure 1. The CT scans of a patient that suffered from post-hepatectomy bilioma (white asterisk) originating from the remnant liver (RL). The left half of the image shows frontal/coronal plain, and the right half of the image shows the remnanttransverse liver (RL). plain. The left half of the image shows frontal/coronal plain, and the right half of the image shows the transverse plain. Table 3. The isolated bacterial strains. Table 3. The isolated bacterial strains. Microorganisms Isolated n % Microorganisms Isolated Bacteroides fragilis 1 6.7 Bacteroides thetaiotaomicronn 1 % 6.7 Bacteroides vulgatus 1 6.7 Bacteroides fragilis 1 6.7 Prevotella buccae 3 20.0 Bacteroides thetaiotaomicron 1 6.7 Prevotella melaninogenica 2 13.3 Bacteroides vulgatus Finegoldia magna 1 3 6.720.0 Prevotella buccae Bilophila wadsworthia3 1 20.06.7 Prevotella melaninogenica Veillonella parvula 2 1 13.36.7 Finegoldia magna Bifidobacterium animalis3 1 20.06.7 Bilophila wadsworthia Anaerobe not otherwise specified1 1 6.76.7 Veillonella parvula Total 1 15 6.7100 Biﬁdobacterium animalis 1 6.7 Anaerobe2.4. Risk not Factor otherwise Analysis speciﬁed for Anaerobic Infection. 1 6.7 Total 15 100

2.4. Risk Factor Analysis for Anaerobic Infection Anaerobic infections were present in 13 out of 245 patients (5%). We performed a stepwise backward binary logistic regression to find risk factors for anaerobic infections in liver resection. As potential predictors, we chose age, BMI, diabetes mellitus 2, preoperative biliary stent, resection and reconstruction of the bile ducts, postoperative bile leak, extent of surgical procedure, American Society of Anesthesiologist (ASA) score, Model of End- Stage Liver Disease (MELD) score, activation of the in-house post-hepatectomy liver failure protocol, malignant disease, and infectious reason for surgery. In the final model, preoperative biliary stent and postoperative bile leak were identified as predictors for anaerobic infections, X2 (3) = 17.017, p 0.001; Cox & Snell 0.072; R2 Nagelkerke R2 0.212. Preoperative stent had an Odds Ratio of 7.790 (CI 95% 2.151–28.221; p 0.002), while bile leak had an Odds Ratio of 5.426 (CI 95% 1.535–19.182; p 0.009).

2.5. Treatment of Anaerobic Infections Of the 13 patients, 7 patients had positive biliary cultures from intraoperative specimens gathered from the bile duct. Six of these patients had a stent, and the remaining were suffering from repeated cholangitis after complicative cholecystectomy and injury of the right hepatic artery. All of these patients had preemptive perioperative coverage of anaerobic bacteria with metronidazole or piperacillin/tazobactam. All of these patients had their bile ducts and or infected segments of the liver resected. None of these patients had recurrent anaerobic infection during their stay. Antibiotics 2021, 10, 139 7 of 12

In the other six patients, the most common sites of origin of anaerobic bacteria were the skin and gastrointestinal tract in 50% and 50% of patients, respectively. In order to resolve the infectious complications, surgical source control was achieved by evacuation of the infected tissues via incision, resection and drainage. None of these complications occurred. Overall, the outcomes were favorable. The overall mortality in the cohort was 6%. The mortality within the group of patients that suffered from anaerobic infections was 15%. However, none of the deaths was attributed to septicemia from anaerobic bacteria.

3. Discussion The role of anaerobic bacteria in infections of the liver is unclear. The most commonly described pathologies are (septic) liver abscesses originating from the biliary tract, gastrointestinal tract, or endocarditis [7,8]. In 1996, Huang et al. reported that up to 25% of liver abscesses grew anaerobic bacteria [8]. The widespread use of percutaneous ablation of liver tumors has also led to an increase in liver abscesses caused iatrogenically [9,10], especially if the patient has undergone previous bilioenteric anastomosis [18,19]. Data on the prevalence of anaerobic bacteria and their role in complications after liver resections are scarce. Surgical site and organ space infections after liver resections pose an increased risk to the patient and significantly prolong hospital stay [20]. We hereby present the first study to focus on the prevalence of anaerobes in microbiological specimen collected during major liver resection and complications following these procedures. We present a series of 245 consecutive patients who underwent major liver resection at our tertiary hepatobiliary center. We were able to identify seven patients with anaerobic bacteria in the bile duct, as well as six patients that developed complications and subsequent anaerobic infections. Unsurprisingly, the most common site of origin of these anaerobes was in the gastrointestinal tract. Therefore, a breach of the natural barrier at the duodenal papilla appears to be a risk factor for anaerobic colonization of the biliary tract, similar to fungal colonization of the bile duct [12]. The number of anaerobic infections appears to increase the longer a stent is in place [12]. Even though a recent meta-analysis suggests that there are less infections with percutaneous drainage of the biliary tract compared to endoscopic drainage, the rates of the bacterial colonization were high in both groups [5,21]. Biliary drainage prior to surgery leads to an increased risk of postoperative infection [22,23]. In a cohort of 475 patients who had undergone preoperative biliary drainage, Sugawara et al. reported that nearly 75% of patients had contaminated bile. Of these patients, 28.6% developed postoperative infectious complications. Bacteriobilia is an independent risk factor of postoperative infectious complications. However, compared to the endoscopic internal drainage, Sugawara et al. did not report a single positive anaerobic culture [5,12]. Perioperative anti-infective treatment in liver surgery is usually done as a single dose prior to incision, with repeated doses based on the length of the procedure and excretion kinetics of the substance used [24]. There are little data that support prolonged anti-infective treatment unless patients are operated in the setting of an infection [24]. Our in-house protocol recommends perioperative antibiotics for a minimum of 3 days in patients who had undergone preoperative stenting or recurrent cholangitis or are at an increased risk of post-hepatectomy liver failure. We frequently cultured bacteria from the bile that were not susceptible to first- and second-generation cephalosporins and thus elected to use third-generation cephalosporins for liver and bile duct surgery. All patients undergoing biliary tract surgery receive perioperative third-generation cephalosporins with metronidazole (and fluconazole in case of a stent). External drainage is very uncommon in our region. Therefore, our perioperative regimen is based on the known local resistance patterns. Whereas there are sufficient data to not recommend prolonged antibiotics in elective and “sterile” liver surgery, there are less data on what to do in the setting of a contaminated bile duct. Sudo et al. found an increased rate of antibiotic resistance in patients that had undergone biliary drainage compared to those who had not. This may in part be due to preoperative antibiotic treatment and stent placement in the setting of cholangitis. Whereas there is a risk of bias, we support their suggestion to take into account Antibiotics 2021, 10, 139 8 of 12

microbiological culture results from specimens gathered from external bile drainage when- ever available [25]. A Japanese randomized controlled trial confirmed that culture-based targeted therapy leads to less surgical site infections [26]. The most frequently reported bacterial species are E. coli and Enterococcus [4,5,23]. There were no reports on anaerobic growth in these trials. We hypothesized that infected biliomas are frequently colonized by anaerobic bacteria. In our study, we were able to culture anaerobic bacteria from two biliomas with polymicrobial infection. The bacterial spectrum of biliomas after liver surgery is considered to be similar to that of cholangitis. A single retrospective analysis of 32 biliomas (most of them after liver surgery) reported the microbiological spectrum of these patients. The authors isolated 121 different bacterial strains from 32 biliomas. Only four of these were anaerobes (all Clostridium perfringens)[4]. A higher rate of up to 10% of biliomas and liver abscesses infected with anaerobes was reported in liver transplant recipients [27,28]. To our knowledge, there is no larger study available that has focused on microbiological data on bilioma after liver resection. Thus, the incidence of anaerobic bacteria in this population is yet to be assessed. Multicentric randomized trials need to be conducted to solve the question of the true prevalence of anaerobic infections in liver resections. Aside from bile and tissue cultures, new but not yet widely available diagnostic tools need to be introduced. A recent study by Dyrhovden et al. from Norway examined next generation sequencing from biliary samples in cholecystitis [11]. They reported up to 23% of anaerobic bacteria in bile samples from gallbladder aspirates and lower numbers from cultures. This suggests that we need to change the way we think about microbiological sampling during surgical procedures. There are no recent studies that have compared tissue and aspirate samples to swabs from infected tissues in abdominal surgery. Most studies comparing different sampling techniques have focused on chronic wounds [29] and infected musculoskeletal tissues [30]. Swabs from infected tissues are still a common specimen sampling in ORs worldwide. Clinicians and microbiologists will need to focus more on the proper collection and work up of samples from abdominal infections. Whereas the bile itself may be inhibitory to the growth of bacteria, we frequently had polymicrobial cultures [31]. The delayed growth of anaerobes and fungi furthermore may lead to therapeutic insecurity, especially when cultures grow bacteria days after surgery that are not covered by empiric antibiotics. Next- generation sequencing (NGS) potentially offers earlier results than cultures. In the near future, NGS may help in the decision process of when to stop empiric antibiotics early, thus preventing resistance formation and limiting unnecessary toxicity and costs. Overall, merely 5% of our patients developed anaerobic infections. Assuming Dyrhov- den’s results could be transferred to major hepatectomies, we are at risk of significant underdiagnosis [11]. In our patient cohort, preoperative biliary stents and postoperative bile leaks were predictors of anaerobic infection. This is similar to our department’s finding that anaerobic bacteria were more frequently isolated in cholecystitis patients who had undergone preoperative endoscopic retrograde cholangiopancreaticography [32]. The 2018 Tokyo Guidelines recommend covering anaerobes in the setting of cholangitis in patients that have biliary obstruction or bilioenteric anastomosis [13]. We therefore continue to em- pirically treat anaerobes in all hepatectomies involving biliary tract resections, preoperative stenting, and postoperative bilioma and bile leaks. Further prospective data are needed to support this approach. The most commonly used antibiotics to treat anaerobes in abdominal infections are metronidazole, piperacillin/tazobactam, and carbapenems [13]. According to a 2020 review by Thabit, all three options have excellent penetration into the peritoneum as well as the biliary tract. However, since there are limited data on the biliary penetration of ertapenem and due to its intrinsic lack of activity against pseudomonas, meropenem or imipenem/cilastatin are likely to be preferred in infections of the liver [33]. Clindamycin, which is perceived to be a good candidate to treat anaerobes in the skin and “above the diaphragm,” has shown decreased activity against gastrointestinal aerobe and anaerobe Antibiotics 2021, 10, 139 9 of 12

pathogens [34] and has fallen out of the general surgeon’s favor for the increased risk of clostridium difﬁcile infection [35] compared to other antibiotics.

Strengths and Limitations To our knowledge, this is the first study to focus solely on the prevalence and clinical presentation of anaerobic bacteria in major liver resection. We presented data from 245 cases of major liver resections from our tertiary hepatobiliary surgery unit. We were able to identify biliary stenting and bile leaks as predictors for anaerobic infections. However, given the overall low prevalence of anaerobes, caution must be taken when generalizing these results. Overall, we were slightly surprised to find only 5% of anaerobic infections over this 3-year period. Given the high rate of cholangiocarcinoma in our cohort and the consecutive increased rate of bile duct obstructions and bile leaks, we expected to find a higher number of positive anaerobic cultures. Due to the retrospective character of this study, we were unable to optimize specimen sampling and were also unable to perform next-generation sequencing to correlate these results to those of our cultures.

4. Materials and Methods 4.1. Data Acquisition We retrospectively screened our hospital information system for all patients who underwent major liver resection at the department of General, Visceral, and Transplan- tation Surgery of the University Hospital of Tuebingen, Germany. All adult patients (age ≥ 16 years) who underwent resection of at least 3 liver segments as an individual procedure at our center between January July 2017 and August 2020 were included in the ﬁnal analysis. The medical reports of these patients were screened for intra- and perioperative microbiological cultures. Cultures were sent at the surgeon’s discretion, most commonly when there were clinical signs of chronic cholangitis or biliary stenting or intraoperative ﬁndings of abscesses or perforation. The study was performed on a consecutive database and was approved by the local ethics committee under the reference number 419/2020/BO.

4.2. Clinical Definitions Major hepatectomy/liver resection is defined as the simultaneous resection of at least 3 segments of the liver. A right hepatectomy is the removal of Coinaud’s segments V-VIII, a left hepatectomy is the removal of Segments II-IV, a right Trisectionectomy is the removal of Segments IV-VIII, and a left Trisectionectomy is the removal of liver segments II-V and VIII according to the 2000 Brisbane terminology [16]. A bilioma is defined as an organized collection of bile, usually with association to the liver resection margin or to the bilioenteric anastomosis that is encapsulated and filled with bile.

4.3. Isolation and Identification of Strains All samples were sent for both aerobic and anaerobic cultures. For the anaerobic cultures, we used a custom-built highly enriched and supplemented Sheep Blood Agar. Therefore, sheep blood (5% defibrinated, Source: Acila) was added to brain-heart infusion agar (Source: Oxoid) and then enriched with IsoVitalex enrichment (Source: BD) at 37 ◦C. The method has been recently described by our microbiology department [36]. The isolates were then identified using Matrix-Assisted Laser Desorption/Ionization-Time Of Flight (MALDI-TOF) mass spectrometry.

4.4. Statistics Comparison between groups was carried out by the Chi-Square test (X2) or Fisher´s exact test (FET) for nominal variables and the Mann–Whitney U-test (MWU) for continuous variables, as appropriate. A probability of less than 0.05 was considered to be statistically signiﬁcant. All p-values reported were the result of 2-sided testing. Where needed, Bonfer- roni correction was applied. Statistical analysis was carried out using IBM SPSS Statistics for Windows, Version 26.0 (IBM Corp., Armonk, NY, USA). Antibiotics 2021, 10, 139 10 of 12

5. Conclusions Anaerobic infections play an unknown rule in the setting of major hepatectomy. This study found an increased rate of anaerobic infection in patients that had undergone preoperative endoscopic biliary stenting. Furthermore, postoperative bile leaks appeared to be a predictor of anaerobic infections. Currently, there are little to no available trials that report the rate of anaerobic infections in hepatobiliary surgery, which may in part be due to the slow and difﬁcult growth rate of anaerobes. Whereas the outcomes of the anaerobic infections in our cohort were favorable overall, all of our postoperative complications caused by anaerobic bacteria had to be treated interventionally and/or surgically in addition to the anti-infective medication. Further data is needed to understand the role of anaerobic bacteria in patients undergoing liver resections and to predict the presence and role of anaerobes in complications after hepatectomy. Newer diagnostic tools, such as next-generation sequencing, may aid in the early and precise identiﬁcation of anaerobic strains.

Author Contributions: Conceptualization, J.S. and A.M.; methodology, J.S., S.B.; formal analysis, J.S., S.B.; data curation, J.S., A.M.; writing—original draft preparation, J.S., A.M.; writing—review and editing, S.N., A.K.; visualization, A.M.; supervision, R.L.; All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the medical faculty of the University of Tuebingen under the protocol code 419/2020/BO; date of final approval 5 June 2020. Informed Consent Statement: Patient consent was waived due to the analysis of pseudonymized data, according to the approved study protocol. Data Availability Statement: The data presented in this study are available on request from the corresponding author. The data are not publicly available du the pseudonymized character of the data. Acknowledgments: We acknowledge support by the Open Access Publishing Fund of the University of Tuebingen. We would like to thank Elke-Henriette Erdei, MSc in Mathematics, PhD Candidate at the Technical University of Dresden, Germany for her support with binary logistic regression and statistics. Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations

ASA American Society of Anesthesiologist CRLM Colorectal Liver Metastasis BMI Body Mass Index CI Conﬁdence Interval FET Fisher´s Exact Test iCCA intrahepatic Cholangiocarcinoma GBCA Gallbladder Carcinoma HCC Hepatocellular Carcinoma MALDI-TOF Matrix-Assisted Laser Desorption/Ionization-Time Of Flight MELD Model of End-Stage Liver Disease MWU Mann-Whitney U-Test NET Neuroendocrine Tumor NGS Next-Generation Sequencing phCCA perihilar Cholangiocarcinoma PHLF Post-Hepatectomy Liver Failure SD Standard Deviation SSI Surgical Site Infection X2 Chi-Square test Antibiotics 2021, 10, 139 11 of 12

References 1. Moreno Elola-Olaso, A.; Davenport, D.L.; Hundley, J.C.; Daily, M.F.; Gedaly, R. Predictors of surgical site infection after liver resection: A multicentre analysis using National Surgical Quality Improvement Program data. HPB (Oxford) 2012, 14, 136–141. [CrossRef][PubMed] 2. Yamashita, Y.-I.; Hamatsu, T.; Rikimaru, T.; Tanaka, S.; Shirabe, K.; Shimada, M.; Sugimachi, K. Bile leakage after hepatic resection. Ann. Surg. 2001, 233, 45–50. [CrossRef][PubMed] 3. Spetzler, V.N.; Schepers, M.; Pinnschmidt, H.O.; Fischer, L.; Nashan, B.; Li, J. The incidence and severity of post-hepatectomy bile leaks is affected by surgical indications, preoperative chemotherapy, and surgical procedures. Hepatobiliary Surg. Nutr. 2019, 8, 101–110. [CrossRef][PubMed] 4. Würstle, S.; Göß, A.; Spinner, C.D.; Huber, W.; Algül, H.; Schlag, C. A retrospective clinical and microbial analysis of 32 patients with bilomas. BMC Gastroenterol. 2019, 19, 50. [CrossRef][PubMed] 5. Sugawara, G.; Ebata, T.; Yokoyama, Y.; Igami, T.; Takahashi, Y.; Takara, D.; Nagino, M. The effect of preoperative biliary drainage on infectious complications after hepatobiliary resection with cholangiojejunostomy. Surgery 2013, 153, 200–210. [CrossRef] [PubMed] 6. Xiong, Y.M.; Rao, X. Clinical and Microbiological Characteristics of Patients with Complicated Intra-abdominal Infections in Intensive Care Unit. Curr. Med. Sci. 2020, 40, 104–109. [CrossRef] 7. Hansen, P.S.; Schønheyder, H.C. Pyogenic hepatic abscess. A 10-year population-based retrospective study. APMIS 1998, 106, 396–402. [CrossRef] 8. Huang, C.J.; Pitt, H.A.; Lipsett, P.A.; Osterman, F.A., Jr.; Lillemoe, K.D.; Cameron, J.L.; Zuidema, G.D. Pyogenic hepatic abscess. Changing trends over 42 years. Ann. Surg. 1996, 223, 600–609. [CrossRef] 9. Su, X.F.; Li, N.; Chen, X.F.; Zhang, L.; Yan, M. Incidence and Risk Factors for Liver Abscess After Thermal Ablation of Liver Neoplasm. Hepat. Mon. 2016, 16, e34588. [CrossRef] 10. Giorgio, A.; Merola, M.G.; Montesarchio, L.; Merola, F.; Gatti, P.; Coppola, C.; Calisti, G. Percutaneous radiofrequency ablation of hepatocellular carcinoma in cirrhosis: Analysis of complications in a single centre over 20 years. Br. J. Radiol. 2017, 90, 20160804. [CrossRef] 11. Dyrhovden, R.; Øvrebø, K.K.; Nordahl, M.V.; Nygaard, R.M.; Ulvestad, E.; Kommedal, Ø. Bacteria and fungi in acute cholecystitis. A prospective study comparing next generation sequencing to culture. J. Infect. 2020, 80, 16–23. [CrossRef][PubMed] 12. Schneider, J.; Hapfelmeier, A.; Fremd, J.; Schenk, P.; Obermeier, A.; Burgkart, R.; Weber, A. Biliary endoprosthesis: A prospective analysis of bacterial colonization and risk factors for sludge formation. PLoS ONE 2014, 9, e110112. 13. Gomi, H.; Solomkin, J.S.; Schlossberg, D.; Okamoto, K.; Takada, T.; Strasberg, S.M.; Ukai, T.; Endo, I.; Iwashita, Y.; Hibi, T.; et al. Tokyo Guidelines 2018: Antimicrobial therapy for acute cholangitis and cholecystitis. J. Hepatobiliary Pancreat. Sci. 2018, 25, 3–16. [CrossRef][PubMed] 14. Capobianco, I.; Rolinger, J.; Nadalin, S. Resection for Klatskin tumors: Technical complexities and results. Transl. Gastroenterol. Hepatol. 2018, 3, 69. [CrossRef][PubMed] 15. Li, J.; Moustafa, M.; Linecker, M.; Lurje, G.; Capobianco, I.; Baumgart, J.; Nadalin, S. ALPPS for Locally Advanced Intra- hepatic Cholangiocarcinoma: Did Aggressive Surgery Lead to the Oncological Beneﬁt? An International Multi-center Study. Ann. Surg. Oncol. 2020, 27, 1372–1384. [CrossRef][PubMed] 16. Strasberg, S.M.; Belghiti, J.; Clavien, P.A.; Gadzijev, E.; Garden, J.O.; Lau, W.Y.; Strong, R.W. The Brisbane 2000 Terminology of Liver Anatomy and Resections. HPB 2000, 2, 333–339. [CrossRef] 17. Bressan, A.K.; Isherwood, S.; Bathe, O.F.; Dixon, E.; Sutherland, F.R.; Ball, C.G. Preoperative Single-Dose Methylprednisolone Prevents Surgical Site Infections After Major Liver Resection: A Randomized Controlled Trial. Ann. Surg. 2020. Publish Ahead of Print. [CrossRef] 18. Welch, B.T.; Schmitz, J.J.; Atwell, T.D.; McGauvran, A.M.; Kurup, A.N.; Callstrom, M.R.; Schmit, G.D. Evaluation of infectious complications following percutaneous liver ablation in patients with bilioenteric anastomoses. Abdom. Radiol. 2017, 42, 1579–1582. [CrossRef] 19. Hoffmann, R.; Rempp, H.; Schmidt, D.; Pereira, P.L.; Claussen, C.D.; Clasen, S. Prolonged antibiotic prophylaxis in patients with bilioenteric anastomosis undergoing percutaneous radiofrequency ablation. J. Vasc. Interv. Radiol. 2012, 23, 545–551. [CrossRef] 20. Kokudo, T.; Uldry, E.; Demartines, N.; Halkic, N. Risk factors for incisional and organ space surgical site infections after liver resection are different. World J. Surg. 2015, 39, 1185–1192. [CrossRef] 21. Rizzo, A.; Ricci, A.D.; Frega, G.; Palloni, A.; De Lorenzo, S.; Abbati, F.; Brandi, G. How to Choose Between Percutaneous Transhepatic and Endoscopic Biliary Drainage in Malignant Obstructive Jaundice: An Updated Systematic Review and Meta- analysis. In Vivo 2020, 34, 1701–1714. [CrossRef][PubMed] 22. Ferrero, A.; Lo Tesoriere, R.; Viganò, L.; Caggiano, L.; Sgotto, E.; Capussotti, L. Preoperative biliary drainage increases infectious complications after hepatectomy for proximal bile duct tumor obstruction. World J. Surg. 2009, 33, 318–325. [CrossRef][PubMed] 23. Hochwald, S.N.; Burke, E.C.; Jarnagin, W.R.; Fong, Y.; Blumgart, L.H. Association of preoperative biliary stenting with increased postoperative infectious complications in proximal cholangiocarcinoma. Arch. Surg. 1999, 134, 261–266. [CrossRef][PubMed] 24. Shinkawa, H.; Tanaka, S.; Takemura, S.; Amano, R.; Kimura, K.; Nishioka, T.; Ito, T.; Miyazaki, T.; Ishihara, A.; Kubo, S. Giv- ing short-term prophylactic antibiotics in patients undergoing open and laparoscopic hepatic resection. Ann. Gastroenterol. Surg. 2019, 3, 506–514. [CrossRef][PubMed] Antibiotics 2021, 10, 139 12 of 12

25. Sudo, T.; Murakami, Y.; Uemura, K.; Hayashidani, Y.; Hashimoto, Y.; Ohge, H.; Sueda, T. Speciﬁc Antibiotic Prophylaxis Based on Bile Cultures Is Required to Prevent Postoperative Infectious Complications in Pancreatoduodenectomy Patients Who Have Undergone Preoperative Biliary Drainage. World J. Surg. 2007, 31, 2230–2235. [CrossRef] 26. Okamura, K.; Tanaka, K.; Miura, T.; Nakanishi, Y.; Noji, T.; Nakamura, T. Randomized controlled trial of perioperative antimicrobial therapy based on the results of preoperative bile cultures in patients undergoing biliary reconstruction. J. Hepatobiliary Pancreat. Sci. 2017, 24, 382–393. [CrossRef] 27. Said, A.; Safdar, N.; Lucey, M.R.; Knechtle, S.J.; D’Alessandro, A.; Musat, A.; Pirsch, J.; Kalayoglu, M.; Maki, D.G. Infected bilomas in liver transplant recipients, incidence, risk factors and implications for prevention. Am. J. Transplant. 2004, 4, 574–582. [CrossRef] 28. Tachopoulou, O.A.; Vogt, D.P.; Henderson, J.M.; Baker, M.; Keys, T.F. Hepatic abscess after liver transplantation: 1990–2000. Transplantation 2003, 75, 79–83. [CrossRef] 29. Copeland-Halperin, L.R.; Kaminsky, A.J.; Bluefeld, N.; Miraliakbari, R. Sample procurement for cultures of infected wounds: A systematic review. J. Wound Care 2016, 25, S4–S10. [CrossRef] 30. Aggarwal, V.K.; Higuera, C.; Deirmengian, G.; Parvizi, J.; Austin, M.S. Swab cultures are not as effective as tissue cultures for diagnosis of periprosthetic joint infection. Clin. Orthop. Relat. Res. 2013, 471, 3196–3203. [CrossRef] 31. Begley, M.; Gahan, C.G.; Hill, C. The interaction between bacteria and bile. FEMS Microbiol. Rev. 2005, 29, 625–651. [CrossRef] [PubMed] 32. Strohäker, J.; Wiegand, L.; Beltzer, C.; Königsrainer, A.; Ladurner, R.; Meier, A. Clinical Presentation and Incidence of Anaerobic Bacteria in Surgically Treated Biliary Tract Infections and Cholecystitis. Antibiotics 2021, 10, 71. [CrossRef][PubMed] 33. Thabit, A.K. Antibiotics in the Biliary Tract: A Review of the Pharmacokinetics and Clinical Outcomes of Antibiotics Penetrating the Bile and Gallbladder Wall. Pharmacotherapy 2020, 40, 672–691. [CrossRef][PubMed] 34. Takesue, Y.; Kusachi, S.; Mikamo, H.; Sato, J.; Watanabe, A.; Kiyota, H.; Yanagihara, K. Antimicrobial susceptibility of common pathogens isolated from postoperative intra-abdominal infections in Japan. J. Infect. Chemother. 2018, 24, 330–340. [CrossRef] [PubMed] 35. Brown, K.A.; Khanafer, N.; Daneman, N.; Fisman, D.N. Meta-analysis of antibiotics and the risk of community-associated Clostridium difﬁcile infection. Antimicrob. Agents Chemother. 2013, 57, 2326–2332. [CrossRef] 36. Buhl, M.; Willmann, M.; Liese, J.; Autenrieth, I.B.; Marschal, M. Prevotella colorans sp. nov., isolated from a human wound. Int. J. Syst. Evol. Microbiol. 2016, 66, 3005–3009. [CrossRef]